Comparing Performance and Quality of Traffic Assignments for Microscopic Simulation

Abstract

Focusing on the tradeoff between accuracy of the assignment and computation time, this paper proposes different traffic assignment methods targeting at microscopic traffic simulation. Moreover, two approaches to reduce the time and the memory consumption required with use of the microscopic simulation soft-ware SUMO (Simulation of Urban MObility) are also proposed and analyzed. The results show that the proposed approaches perform better than the original one without compromising the algorithm quality. Regarding the performance analysis among different traffic assignments the results indicate that the saving on compu-tation time is significant with use of macroscopic assignments. Also, the computa-tion time used for some microscopic methods (e.g. the one-shot method) is com-petitive with that of the macroscopic assignments. While the exact parameterization as well as the sensitivity of the methods to the size of the scenarios still need further investigation, it seems favorable to employ microscopic assignment techniques or hybrid methods for producing a good traffic assignment for a microscopic simulation.